finest 0.0.4

FineST: Fine-grained Spatial Transcriptomic

A tatistical model and toolbox to identify the super-resolved ligand-receptor interaction with spatial co-expression (i.e., spatial association). Uniquely, FineST can distinguish co-expressed ligand-receptor pairs (LR pairs) from spatially separating pairs at sub-spot level or single-cell level, and identify the super-resolved ligand-receptor interaction (LRI).

It comprises three components (Training-Imputation-Discovery) after HE image feature is extracted:

• Step0: HE image feature extraction

• Step1: Training FineST on the within spots

• Step2: Super-resolution spatial RNA-seq imputation

• Step3: Fine-grained LR pair and CCC pattern discovery

Installation

FineST is available through PyPI. To install, type the following command line and add -U for updates:

pip install -U FineST

Alternatively, install from this GitHub repository for latest (often development) version (time: < 1 min):

pip install -U git+https://github.com/LingyuLi-math/FineST

Installation using Conda

$ git clone https://github.com/LingyuLi-math/FineST.git
$ conda create --name FineST python=3.8
$ conda activate FineST
$ cd FineST
$ pip install -r requirements.txt

Typically installation is completed within a few minutes. Then install pytorch, refer to pytorch installation.

$ conda install pytorch=1.7.1 torchvision torchaudio cudatoolkit=11.0 -c pytorch

Verify the installation using the following command:

python
>>> import torch
>>> print(torch.__version__)
>>> print(torch.cuda.is_available())

Get Started for Visium HD data

Illustrate using a single slice of 10x Visium HD human colorectal cancer (CRC) data with 16-um bin.

Step0: HE image feature extraction

Input

• Visium_HD_Human_Colon_Cancer_tissue_image.btf: Raw histology image (.btf Visium HD or .tif Visium)

• tissue_positions.parquet: Spot/bin locations (.parquet Visium HD or .csv Visium)

Output

• HD_CRC_16um_pth_32_16_image: Segmeted histology image patches (.png)

• HD_CRC_16um_pth_32_16: Extracted image feature embeddiings for each patche (.pth)

python .FineST/HIPT_image_feature_extract.py --dataset HD_CRC_16um --position ./Colon_Cancer/square_016um/spatial/tissue_positions.parquet --image ./Colon_Cancer/Visium_HD_Human_Colon_Cancer_tissue_image.btf --output_path_img ./HD_CRC_16um_pth_32_16_image --output_path_pth ./HD_CRC_16um_pth_32_16 --patch_size 32 --logging_folder ./Finetune/HIPT_image_feature_extract/

HIPT_image_feature_extract.py also output the execution time:

• The image segment execution time for the loop is: 62.491 seconds

• The image feature extract time for the loop is: 1717.818 seconds

Step1: Training FineST on the within spots

Step2: Super-resolution spatial RNA-seq imputation

Step3: Fine-grained LR pair and CCC pattern discovery

Detailed Manual

The full manual is at finest-rtd-tutorial for installation, tutorials and examples.

• Sub-spot level prediction by FineST (For Visium dataset).

• Super-resolved ligand-receptor interavtion discovery by FineST.